The big idea
Governments and humanitarian groups can use machine learning algorithms and mobile phone data to provide help to those who need it most during a humanitarian crisis, we have found in new research.
As the COVID-19 pandemic spread in early 2020, our research team helped Togo Ministry of Digital Economy and Givedirectlya non-profit organization that sends money to people living in poverty, are turning this idea into a new kind of aid program. simple idea behind this approach, as we explained in the review Nature March 16, 2022, is that the rich use the telephone differently from the poor. Their phone calls and text messages follow different patterns and they use different data plans, for example. Machine learning algorithms, which are sophisticated pattern recognition tools, can be trained to recognize these differences and infer whether a given mobile subscriber is rich or poor.
First, we collected recent, reliable and representative data. Working on the ground with partners in Togo, we conducted 15,000 telephone surveys to collect information on the living conditions of each household. After comparing survey responses with data from mobile phone companies, we trained the machine learning algorithms to recognize phone usage patterns characteristic of people living on less than $1.25 a day.
The next challenge was to determine if a system based on machine learning and phone data would be effective in getting money to the poorest people in the country. Our assessment indicated that this new approach worked better than the other options considered by the Togolese government.
For example, focusing entirely on the poorest townships – which are analogous to US counties – would have provided benefits to only 33% of people living on less than US$1.25 a day. In contrast, the machine learning approach targeted 47% of this population.
We then partnered with the Togolese government, GiveDirectly, and community leaders to design and pilot a cash transfer program based on this technology. In November 2020, the first beneficiaries were registered and paid. To date, the program has provided nearly $10 million to approximately 137,000 of the country’s poorest citizens.
why is it important
Our work shows that data collected by mobile phone companies, when analyzed with machine learning technology, can help direct help to those who need it most.
Even before the pandemic, more than half of West African nation 8.6 million people lived below the international poverty line. As COVID-19 further slowed economic activity, our surveys indicated that 54% of all Togolese were forced to skip meals each week.
The situation in Togo was not unique. The slowdown resulting from the COVID-19 pandemic plunged millions of people into extreme poverty. In response, governments and charities have launched several thousand new aid programs, providing benefits to more than 1.5 billion people and families around the world.
But in the midst of a humanitarian crisis, governments are struggling to determine who is most in need of urgent help. In ideal circumstances, these decisions would be based on in-depth household surveys. But there was no way to collect this information in the midst of a pandemic.
Our work helps demonstrate how new sources of big data, such as information collected from satellites and mobile phone networks, can help target aid in crisis conditions when more traditional data sources are not available. available.
We conduct follow-up research to assess the impact of cash transfers on recipients. Previous results indicate that cash transfers can help increase food security and improve psychological well-being in normal times. We assess whether this aid has similar results in times of crisis.
It is also essential to find ways to register and pay people without telephones. In Togo, about 85% of households had at least one telephone, and phones are frequently shared within families and communities. However, it is unclear how many people who needed humanitarian assistance in Togo did not receive it due to their lack of access to a mobile device.
In the future, systems that combine new methods leveraging machine learning and big data with traditional survey-based approaches are expected to improve the targeting of humanitarian aid.
Emilie AikenPhD student in Information, University of California, Berkeley and Joshua Blumenstock, associate professor of information; Co-director of the Center for Effective Global Action, University of California, Berkeley